Cross Repo Guards — AI skill for Claude Code
Cross-repo guards for AI coding agents.
How to install Cross Repo Guards
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open yybd/cross-repo-guards and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Cross Repo Guards does
Cross-repo guards for AI coding agents. Refuses a blind write or a sweeping commit into a repo the session is not standing in, and closes every place it left open. A Claude Code plugin marketplace: grove.
Alternatives in AI
README
*English · [עברית](README.he.md)*
Grove (Cross-Repo Guards)
**The safety and orchestration layer for AI agents in multi-repo environments.**
An AI agent (like Claude Code) is a brilliant tool when operating within a single project folder (Repository). However, in the real world, tasks cross boundaries: shipping a feature requires a code change (App), a data schema update (Hub), and editing a marketing page (Web). The moment the agent "reaches out" to a foreign directory, it loses touch with reality. Because the `git status` command is limited to its current working directory, the agent might blindly overwrite another developer's open files, push code to the wrong branch, or perform a destructive blind commit.
**Grove was built to solve exactly this.** It is a set of guardrails that run inside the session, at the moment of the tool call: they detect the anomaly, refuse the two actions that cannot be undone, and hand the agent back the picture it was missing so it can correct itself.
The Components: What does Grove do?
The system provides 3 critical solutions:
1. Active Guardrails (The Guard Hooks)
- Blocking Blind Commits (
scope_foreign_commits): If the AI stands in directory A and attempts to rungit commit -ain directory B, the system blocks it. Instead of a generic failure, it injects the true state of the foreign directory (including the branch name and open files), instructing the AI to commit only specific files. - Honoring Boundaries (
honor_ownership): If a file is claimed by another project (via.claude/owns.json), the hook refuses to let the AI edit it, redirecting the agent to the proper location.
2. Smart Contextual Committing (`close.py`)
An agent updating 3 projects simultaneously leaves a trail of changes. Standard Git commands break down here. The script `$GROVE/tools/close.py --commit` reads which files **this specific session** wrote across every directory, and stages only those — never a sweep. What it claim
Related Skills
Repo Guards
RepoGuards — cross-repo guards for AI coding agents. Refuses a blind write or a sweeping commit into a repo th
Mod Squad
Claude Code mods: small plugins for live panes, cost-aware model routing, and safety guards. Install from the
Neo Agent Skills
Install the workflow substrate a cross-model AI team actually runs on: ticket intake, PR review gates, lane co
Cross Agent Session Resumer
Resume AI coding sessions across providers: converts Codex, Claude, Gemini, and other session formats through
Persona Portability Benchmark
One persona, one frozen memory, N models: how much of an agent's character survives a model swap? Harness + bl
Karajan Code
Your AI writes the code; Karajan governs how it happens: TDD-first method, cross-AI review on every commit and
Related Agents
FirstTimeUser
Code-blind first-time customer. Given only a URL and a goal, uses the app in a browser as a stranger would and
Drafting Agent
Template-filling agent for any stage. Use when one named template needs a first complete draft from supplied e
Distill Critic
Adversarial checker for the distill skill. Given a reply's substance, its distilled version, and the proposed